search

linkedtales / scrapedin

LinkedIn Scraper (currently working 2020)
Apache License 2.0
599 stars 174 forks source link

Any reason to continue using puppeteer: 1.13.0 ? #135

Open acanimal opened 4 years ago

acanimal commented 4 years ago

Puppeteer is at version 5. Why is the project continue depending on the version 1?

I currently found that --lang values passes as the puppeteerArgs are ignored and seems it is a chromium bug.

Any possibility to upgrade the version? At least until version 3 which requires node 10

leonardiwagner commented 4 years ago

Hi @acanimal , actually only due to former stability.. it's working.. so 🤷 , however if you have a reason to update, let's do it!

We can update to version 3 if it's the latest to support node 10, no problems. Pull requests are appreciated.

acanimal commented 4 years ago

Lets try to get some free time this weekend 🚀

neverton1970 commented 4 years ago

Hi, I noticed the scraper stopped working in the last few days, but we use a proxy service and the browser was crashing as soon as it was opened. I wasn't sure if it because we're using authentication or not and puppeteer 1.13 didn't like an update.

On a pure guess, I've updated puppeteer 1.13 to 5.1.0 and it's now working fine, with authentication so my current working theory is that the proxy provider ( can't remember their name off the top of my head) have possibly updated the authentication and the old 1.13 version can't handle it. Upgrading it to 5.1.0 appears to be working fine. 

Obviously not a full regression test, but the 10-15 accounts I've tried to scrape so far, have worked fine under 5.1.0